Creating a "personalized" cancer therapy strategy requires decoding the genome sequence of the individual's tumor for genetic errors, also known as mutations. In essence, this is like troubleshooting the "operating system" of a computer by searching for errors by scanning through every line of code that may be relevant. In the case of cancer, the relevant errors may influence drug response, tumor growth and ultimately prognosis. Generally, this has been difficult to accomplish because tumors stored in paraffin, the preferred method for all clinical laboratories, are extremely difficult to work with using the methods of current molecular analysis. Therefore, the development of novel molecular diagnostics which could help patients has been extraordinarily limited. We are developing an approach which will enable highly sophisticated ways of determining which mutations contribute to cancer development, its aggressiveness and its response to therapy. By making use of novel molecular assays integrated with what is commonly referred to as "next generation" sequencing, we are developing technologies which will enable any research group to conduct large scale analysis of cancers at a fraction of the time and cost currently required using archival material. In addition, these technologies will enable much more sensitive detection of mutations than is feasible with current sequencing methods. The radically improved approaches we are developing will substantially accelerate the identification of mutations which contribute to prognosis (e.g. survival, recurrence of cancer) and prediction (e.g. response to specific targeted therapies). Thus, large scale studies of thousands of individuals can now be accomplished using archival material which is commonly available and associated with critical clinical information that will lead to improved DNA diagnostics. PUBLIC HEALTH RELEVANCE: We are developing an approach which will radically improve and expand our ability to detect genetic errors in mutations from archival tumor material. Current methods are expensive, difficult to conduct, inefficient, lack sensitivity and are limited to extracting only small portions of genetic information from tumors. Our approach will dramatically increase any scientist's ability to scan specific DNA sequences for the genetic errors which influence cancer patient prognosis and their response to specific targeted therapies.